The 7 Balance Points of engines in a Full Career Mode.

[Rakaydos]

There has recently been a lot of discussion about the SLS engines in 23.5... and while this topic is tangental to the topic, I do NOT want it to turn into a "Engines are OP!" "Then dont use them" argument. We've already got, like, 3 or 4 threads for that. :/ (And as a personal note, I suggest everyone try the Stock Engine Rebalance mod by Stupidchris before arguing. Useful RM55 radial engines are useful)

This topic is specifically addressing the concern that "Budgets will balance them." I am personally of the opinion that the game should be balanced for sandbox- I will attempt to show that a sandbox-balanced game can also be balanced in a Tech and Budget limited carear.

In sandbox, I've pointed out 4 "balance points," and 3 of them relate to how much thrust an engine has.

Part Effectiveness (or just "Power") is how much thrust you can get from a single part. Seems straight foreward enough, until you start running into quad-coupled LV engines having better performance than their larger cousins. It also specifically doesnt care how much it weighs, as that is the next point.

Thrust to Weight Ratio is a term you see fairly often on these forums, but for engines specifically the TWR defines an absolute upper boundry to where they can land or take off from. It also gets used as a shorthand for other aspects, but it ist to be confused with...

Thrust to Cross Section ratio. Basically, how tall can you make a rocket who's entire bottom is made of that engine? The LV45 has less power than a skipper, but it has a higher thrust by cross section and a better ISP, which is what makes LV45 and LV30 clusters so popular.

ISP is the last sandbox point- how much fuel do you need to carry to get where you need to go.

In carear mode, the tech tree is considered a balance point by some- a common defence of the SLS parts seems to be "but they're at the end of the tech tree," and that real life technological progress is a justification to obsolete earlier engine designs. I disagree that carear mode tech can justify overpowered parts, but the tech tree is still something to pay attention to... as a tutorial. As said by a developer...

November 13, 2013 12:16:04 PM MST

The tech tree was always meant to work as an introduction tool to KSP. A more 'conventional' tutorial would have probably worked too, but it would be exceedingly long and pretty boring. Plus, I think a game should always give players means to discover it by simply playing, and this is what we set out to do with R&D.

I don't think any player, veteran or first-timer, would enjoy sitting through an overly long 'lesson' on what each part does. And more importantly, they wouldn't get to experience first hand the problems that some parts exist to solve. If you've never experienced the problem of tall ships falling over on the pad, then you won't fully appreciate the fact that launch clamps exist.

In the end, the R&D system works to let everyone experience the process of gradually developing your space program, learning as you go how to overcome the problems you encounter. For new players, this allows them to get into the game without having to sit through an introduction, and for veterans, the restrictions are also a good thing, as they set them (or rather 'us') up with new challenges that require us to come up with new ways to tackle old problems.

The tech tree is meant to balance Expected Player skill. That's why the small, well rounded LV engines are presented to the new player first, while additional complications like air breathing engines are delayed. The aerospike is rarely worth using outside an endgame eve surface return or Jool datum return, so it shouldn't clutter up a newbie's engine tab until their tech tree is complete, despite it's otherwise lackluster performance.

Which leaves the most speculative balance point, Cost. There's been prices in the VAB since forever, but noone has any idea how budgets will work.

My theory for Budgets is that they will be a limit-per-launch, but not carry over between launches. (beyond success bonuses/failure penalties) If this is correct, Price would be balanced against Part Effectiveness, from the sandbox balance points. A skipper is more expensive than a LV45, but cheaper than 4 LV45s and a rockomax-to-4-engine adapter. An SLS engine would be even more expensive, but not to the point that it's cheaper to make an asparagused monster to get the same payload to orbit. Having Cost reinforce Part efficiency would balance various designs by capability- an asparagused monster for X budget has similar capability to a powerful cluster of high tech parts for the same budget- and if you dont have enough budget, getting beyond kerbin SoI may be beyond your capability, even with a fully unlocked tech tree. This is the very problem NASA has in real life.

I could be wrong about budgets- but noone else knows either. to claim some unknown system will balance a part has to be some kind of logical falacy. Appeal to... something. I dunno, it's late and I'm rambling. Let me know what you think.

[Dodgey]

This is a very well thought out concept which I hope will end a few discussions. One question is could you please explain the cross section ratio a bit more?

[Levelord]

That's actually an interesting thought. While the new parts are more efficient and powerful than the older set of engines, this would likely be offset by how expensive they would be when budgets get introduced. Making the Mainsail a slightly inferior, but vastly cheaper alternative.

[maccollo]

That's actually an interesting thought. While the new parts are more efficient and powerful than the older set of engines, this would likely be offset by how expensive they would be when budgets get introduced. Making the Mainsail a slightly inferior, but vastly cheaper alternative.

I wouldn't mind them perhaps being slightly more efficient, but they are not "slightly more efficient", but almost twice as good.

They are so much better I was able to get a 4.4% mass fraction to orbit...RSS. I can beat the Saturn 5, the most efficient rocket that has ever flown, with stock engines and stock fuel fractions:confused:

The highest mass fraction I never got before was just over 2%.

Now it if was... like 20% better I think it would be a bit more reasonable.

Edited April 7, 2014 by maccollo

[Jouni]

They are so much better I was able to get a 4.4% mass fraction to orbit...RSS. I can beat the Saturn 5, the most efficient rocket that has ever flown, with stock engines and stock fuel fractions:confused:

The highest mass fraction I never got before was just over 2%.

Now it if was... like 20% better I think it would be a bit more reasonable.

The real SLS is supposed to have even more efficient engines. The Space Shuttle also had a higher payload fraction than the Saturn V, if you include the orbiter (except for the main engines etc.) in payload.

It's not really meaningful to say that some engines are 20% better than others. Surely, if you set a certain delta-v target, one set of engines can lift 20% more payload than another set of engines to an orbit that requires that much delta-v. But if you change the delta-v target, the difference in payload capacity also changes.

[Javster]

You know how someone made all these really short videos on every block in Minecraft? Someone should make them for KSP, although they vary a lot less.

[Wayfare]

The cause of most of these discussions appears to me a misunderstanding of where KSP is right now. I believe KSP isn't ready to be balanced yet. If we can tell anything from the current state of the game, it's that balance has been a low priority on Squad's checklist so far. And that makes sense. Squad has been building out core gameplay mechanics and they're not done with that. Trying to achieve real balance in the game at this point would be a waste of effort.

The fact that the SLS parts are deliberately better than the older stock parts is actually a sign that balance is starting to weigh into the equation more heavily now. With a contract system on the horizon, and a budget being the natural progression of that, KSP is getting to the point where it makes sense to have a gradient where later, more expensive parts are objectively better than older ones. Make no mistake: balancing KSP will be a massive effort. Metrics like the one suggested by Rakaydos here will help a lot. Community feedback will too, as gaming communities are full of number crunchers and min-maxers who are more than ready to provide hard data for balancing purposes. That is not an insult. I believe such people are vital in ferreting out the imbalances that will invariably creep into even the best-laid plans.

One thing I'm not sure of is whether Squad has anyone employed who has that single-minded "I don't care how you intended this part to work"-mentality to lead that effort. But if they don't, Squad has shown itself quite ready to hire individuals with specific talents to further some stage of KSP's development. Of course that talent might be hiding inside the current dev pool for all we know - Squad simply hasn't worked on that yet.

[Thogapotomus]

I guess I fail to see where the cost issue is balanced in Sandbox? I posted a comment at the tail end of another one of these discussions that seems relevant, so I'll just leave it here.

"At first I thought they were fine due to tech progression in career mode. However after reading through these posts I've changed my stance. I'm inclined to agree that each engine in the game should have it's respective place on the ISP curve and all should strive to stick to it.

I do want tech progression though. So the solution for me would be to allow upgrades in the R&D Department. Instead of new parts outclassing old, your science and money could be spent to raise the ISP for all engines across the board. Heck, you could have multiple upgrades for different things like engine thrust, electricity consumption/storage/generation, ASAS torque, RCS ISP, lighter materials, etc. A good one I think would be structural integrity which could go from .23 strengths up to .23.5 strengths.

All in all, there has to be some tech progression in career for it to be enjoyable to me. A system like this would still balance all engines while having better payload fractions at the end of a completed career."

[Rakaydos]

I guess I fail to see where the cost issue is balanced in Sandbox? I posted a comment at the tail end of another one of these discussions that seems relevant, so I'll just leave it here.

If I were doing the balancing, In carear, cost would slightly bias toward larger parts over collections and clusters of smaller parts, but not actually be used as a performance balancer. So with budgets enabled, you cant make a Tech Zero SRB stack that does a Minmus landing as your very first mission, but you can easilly make a basic rocket that goes up and deployes a parashute. Successful missions give you a bigger budget, and by the time a new player is ready to fly to minmus, you dont NEED a stack of SRBs a hundred meters high.

[Pecan]

Nice post, nicely thought out. Travert's mass-optimal engne charts indicate though that far from being 'lacklustre' the aerospike is often the launch engine of choice (for mass optimisation).

[xrayfishx]

The cause of most of these discussions appears to me a misunderstanding of where KSP is right now. I believe KSP isn't ready to be balanced yet. If we can tell anything from the current state of the game, it's that balance has been a low priority on Squad's checklist so far. And that makes sense. Squad has been building out core gameplay mechanics and they're not done with that. Trying to achieve real balance in the game at this point would be a waste of effort.

If they keep adding things without balancing them every once in a while, they will still be adding things to fill in the gaps that exist now so if they then decide to balance it later, they will find that they have many things that shouldn't really be there and still lack some important thing and then they have to work around that and won't be able to reach a good balance. If every once in a while they did a balance pass, they would find what's missing and make future updates with this in mind. Sure, every large update like career or money will mean a lot of rebalancing, but a whole lot less than if they don't balance anything until before the final release.

[Rakaydos]

Unless repeatedly spawnig capsules to rack up "unspent budget" is their idea of a good time, I doubt the developers will add bugetary rollovers. And unless timewarping for a few years doing nothig is also appealing, I doubt there will a "money over time" system.

Without them, there is no reason to make a vessel "smaller" than your budget will allow. (or if you do, sell your excess budget for Reputation and see if there's any other missions you can complete at the same time for more money!) Therefore, the Cost/effectiveness ratio of parts will deturmine what is best used in carear. (for instance, I doubt we'll see many Orange tanks, if the half-sized fuel cans are still a quarter of the price for half the fuel)

The best approach, IMO, to pricing parts is to reinforce part efficency. It should be cheaper to buy an orange tank than 2 oildrum fuel cans, for instance, but a single oildrum fuel can should still be significantly cheaper than an orange tank, if you dont need the extra fuel.

This approch, however, means OP parts are still OP. Conversly, balancing for sandbox will remain balanced for carear.

[Jasel]

I myself am hoping for a type of "quest" system (in this case they would be missions) where you are graded on budget and efficiency for receiving a mission to return a sample return from Eve or put a science lab/telescope in high Kerbin orbit. This way it doesn't matter when you get the tech, it's how you use it.

[Tweeker]

I have noticed that I rarely use the poodle, I hadn't really given this much thought until ARM came out. Then I started a new career mode, and noticed that the LFB KR-1x2 effectively eclipses the mainsail. This discussion seems to be well under way on it's own. While thinking about this I got sidetracked thinking about the poodle. The poodle's real problem is it's low TWR. It is currently in the bottom 1/4 @ a TWR of 9.0

In most any application any other engine is a better choice. 2 Mk 55's make an excellent substitute having better thrust, & lower weight. The problem is made worse by being right in the middle of the 200-250KN pack.

to really set the poodle apart it needs to be buffed to:

550Kn and 3.5 tons.

This puts it right in the middle of the pack as far as TWR @ 16.01, and the thrust is in between 2 & 3 LVT-30 or LVT-45

I think this would put it far too close to the skipper's envelope, however, so the Skipper would need an adjustment to:

900KN and 5.5 tons.

this would keep the skipper's TWR the same, and place it in between the new poodle & mainsail.

Or you could adjust the skipper to:

1000 KN and 5 tons.

This would have the same benefit of moving it's envelope away from the poodle, and also place it's TWR in between the new poodle and mainsail. Giving a nice ramp up as far as the 2.5M rockomax engines are concerned.

poodle 16.01

skipper 20.38

mainsail 25.5

[Tweeker]

In response to this someone said:

The Poodle is supposed to be an efficient low-tech vacuum engine, filling the niche where you need more thrust than the LV-909 provides, with a better efficiency than the LV-T30 and the LV-T45. Its main problem is its low TWR, which makes it worse than any of its competitors.

To fix the situation, we need two changes. First, reduce the mass of the Poodle to 2 tonnes, making it a bit better than a cluster of four LV-909s. When big engines are available, they should be a better choice than a cluster of smaller engines from the same class. Then, reduce the Isp of the LV-T30 and the LV-T45 to 300/350 seconds (to match the Skipper). Those engines are supposed to be powerful first-stage engines, not efficient vacuum engines.

I think that to make the poodle stand out it's thrust needs to be in a different range than the lvt 30/45 something above 300 at least. otherwise you are better off using the lvt 30 or 45, as they have the same thrust, more or less and a better TWR.

As far as the ISPs of the rockomax engines I think they could do a better job differentiating the engine's rolls.

as it stands the ISPs are:

Poodle 270 390

Skipper 300 350

Mainsail 280 330

This is fairly good, but I think it would be better if the ISPs where

Poodle 270 390

Skipper 285 360

Mainsail 300 330

This makes the Mainsail obviously the best choice at sea level, the mainsail takes over at mid altitudes, and is worse at sea level than the Mainsail ate sea level, and the Poodle in vacuum. The Poodle stands out as the best choice in vacuum.

[Rakaydos]

I have noticed that I rarely use the poodle, I hadn't really given this much thought until ARM came out. Then I started a new career mode, and noticed that the LFB KR-1x2 effectively eclipses the mainsail. This discussion seems to be well under way on it's own. While thinking about this I got sidetracked thinking about the poodle. The poodle's real problem is it's low TWR. It is currently in the bottom 1/4 @ a TWR of 9.0

In most any application any other engine is a better choice. 2 Mk 55's make an excellent substitute having better thrust, & lower weight. The problem is made worse by being right in the middle of the 200-250KN pack.

to really set the poodle apart it needs to be buffed to:

550Kn and 3.5 tons.

This puts it right in the middle of the pack as far as TWR @ 16.01, and the thrust is in between 2 & 3 LVT-30 or LVT-45

I think this would put it far too close to the skipper's envelope, however, so the Skipper would need an adjustment to:

900KN and 5.5 tons.

this would keep the skipper's TWR the same, and place it in between the new poodle & mainsail.

Or you could adjust the skipper to:

1000 KN and 5 tons.

This would have the same benefit of moving it's envelope away from the poodle, and also place it's TWR in between the new poodle and mainsail. Giving a nice ramp up as far as the 2.5M rockomax engines are concerned.

poodle 16.01

skipper 20.38

mainsail 25.5

Keep in mind that thrust isnt the only balancing factor you can play with. Keeping the poodle's thrust (and, compared to the size 2 engines, thrust by cross section) low, you can play with weight to affect its TWR, and it's ISP for fuel efficency.

Really, Thrust and Thrust by cross section are the same value, when compared to same size engines. Cross section only matters when you start comparing clusters to single larger engines, with the cluster being less part efficent (but should probably be slightly more effective to make up for it, and more expensive in carear)

Also...

I put together this rocket in Stock Rebalance mod, but you could probably build one similar-starting with the smallest engine and just enough fuel to give it a 1.4 to 1.5 TWR overall, put it on top of the next smallest engine that ccan still get that TWR, then the next, and so on. (with the rework, adapters have fuel, so thaat's not just aerodynamics)

Pay attention to the size of each stage- the lv909, for instance, has only an adapter fueling it, but the RM48-7s has a stack of Oscar Bs taller than an orange tank. (and making flying this thing damn near impossible) I thought this was interesting.

[Rakaydos]

This is fairly good, but I think it would be better if the ISPs where

Poodle 270 390

Skipper 285 360

Mainsail 300 330

This makes the Mainsail obviously the best choice at sea level, the mainsail takes over at mid altitudes, and is worse at sea level than the Mainsail ate sea level, and the Poodle in vacuum. The Poodle stands out as the best choice in vacuum.

In the mod I favor, the poodle has been buffed to have an ISP of 410 in vaccum. The mainsail also got shifted to being a high thrust upper stage, between the Skipper and it's old stats, to make room for the LFB using it's old stats.

I think there needs to be a low thrust, 600 ISP engine to fit between the lander engines like the poodle and 909, and the NERVA and Ion. The ant engine would be a good canidate IMO, but others disagree.

[Tweeker]

I would also like to see an engine that slots inbetween the upper end of the rocket engines and the nuclear engine maybe an early version of the nuclear, with ISP in the 600 range and the LV-N pushed up the tech tree one tier.

[Tweeker]

Keep in mind that thrust isnt the only balancing factor you can play with. Keeping the poodle's thrust (and, compared to the size 2 engines, thrust by cross section) low, you can play with weight to affect its TWR, and it's ISP for fuel efficency

I had originally thought about this. If you drop the poodle to 1.5t you end up with a TWR of 14.95, this makes it very similar to the LVT-30 & -45

lvt-30 1.25 215 17.5

lvt-45 1.5 200 13.6

poodle 1.5 220 14.95

I think that by raising the thrust you keep the poodle from competing with these engines, otherwise it is a better choice to use an adapter and a 1.25m engine. Honestly this is what keeps me from using it most of the time. Many times I've built a rocket and find that the last stage before orbit needs alittle more punch than the poodle offers. so I end up with a cluster of MK 55s or 2 lvt-30s on the side.

[Jouni]

If we look at real rocket engines, we see two major factors: fuel and role. Engines using kerosene are not very efficient, but because kerosene is cheap and easy to handle, it's more feasible to build big powerful engines using it. Engines burning hydrogen are more efficient but also more expensive to use. Lower stage engines are designed to be more powerful and achieve better Isp at sea level. Upper stage engines don't necessarily have any better vacuum Isp, but they are cheaper and more simple.

We don't have money in KSP at the moment, so we can't really have kerosene-burning upper stage engines. The other three categories remain feasible:

• Kerosene-burning lower stage engines, such as the Mainsail. We don't have different fuels in the stock game, but we can simulate the cheap fuel by having these engines produce more thrust for their size. Isp should be low both at sea level and in vacuum.
  
• Lower stage engines using hydrogen as their fuel, such as the Skipper, the LV-T30, and the LV-T45. Realistically, these should have good Isp both at sea level and in vacuum, at the price of being more expensive for their thrust than upper stage engines. Because costs are not a factor in the game at the moment, we should reduce the vacuum Isp of these engines to make upper stage engines more useful. As lower stage engines, they should have a good TWR, but produce less thrust for their size than kerosene-burning engines.
  
• Upper stage engines, such as the LV-909 and the Poodle. These should have poor atmospheric Isp but good vacuum Isp. (Realistically, these engines might have a slightly worse vacuum Isp than lower stage engines, which they make up with simplicity and low cost.) Thrust and TWR are not as important as for lower stage engines.
  

When it comes to engines, bigger is usually better. Big engines can often be made more efficient than smaller engines. As long as rocket engines remain a niche product, economies of scale favor clusters of small engines, which can produce the same thrust for less price. On the other hand, if rocket engines are ever mass produced in the way jet engines are, big engines can also become more cost-effective.

Based on these principles, I would say that the Mainsail, the Skipper, and the LV-909 engine are fine. The LV-T30 and the LV-T45 should have their Isp values reduced to match the Skipper, in order to make the upper stage engines more useful. The Poodle should be slightly better than a cluster of LV-909s, which can be achieved by reducing its mass to 2 tonnes. I'm not sure what I want to do to the tiny engines and the radial engines.

Edited April 13, 2014 by Jouni

[Rakaydos]

I would also like to see an engine that slots inbetween the upper end of the rocket engines and the nuclear engine maybe an early version of the nuclear, with ISP in the 600 range and the LV-N pushed up the tech tree one tier.

I wouldnt step on the nerva's toes quite like that. One of the reasons I suggest the Lv1 and LV1R for the slot is because they're TINY, but have radial mounting so you can scale them up by spamming them.

So the ion engines are the most efficent space engine, but have the lowest TWR, and have power management concerns.

The NERVA has good thrust and the second highest ISP, but is bulky for its thrust and can be difficult to build around.

This "Vasmir Ant" would have a high ISP, and it's tiny thrust is matched to its tiny size, so you could put a dozen around even a 1m tank to get whatever thrust you need without fiddling with cubic struts... but with a poor TWR, the more you add the less delta V you get.

Then you'd have the Poodle and 909, and then the rest of the engines.

[Tweeker]

I don't really think that it would step on the NERVA's toes to add an early version, with smaller thrust, and lower ISP, it would just show good advancement. I don't think that pushing it up the tech tree one tier is unreasonable, either as it is pretty much "the keys to the kingdom" as far as being able to get anywhere you want in the solar system.

[Comrade Jenkens]

I think that the NERVA should be pushed to the end of the tech tree and given a buff in thrust. In real life they produce a similar amount to chemical rockets.

Before I was happy with it being low thrust due to sandbox balance but now we have efficient, light and powerful chemical engines there isn't a reason to keep it such low thrust any more.

[Rakaydos]

I don't really think that it would step on the NERVA's toes to add an early version, with smaller thrust, and lower ISP, it would just show good advancement. I don't think that pushing it up the tech tree one tier is unreasonable, either as it is pretty much "the keys to the kingdom" as far as being able to get anywhere you want in the solar system.

Ah, you are going for the "inferior model to be replaced" theory. I disagree with that school of thought. What I want is to unlock Different engines, at different points on the Power/weight/efficency matrix, such that when the tech tree is fully unlocked, the LV30 at the beginning is still a good choice for most things, but not as good at any area as the specialized engines.

[Tweeker]

Maybe not so much a model that is made obsolete, but a nerfed version of each engine, and the normal one. or if you want to get fancy a nerfed, an normal and a buffed version, with sandbox having the "full" version.